ami2py

<p align="center"> <img src="logo.png" alt="ami2py logo" width="50%" height="50%" /> </p>

Index

• Short Introduction
• Python Package API
• Examples
• RUST Backend
• ami_cli
• Examples
• Architecture Overview
• FAQs

Short Introduction

Python package for reading and writing AmiBroker databases. The binary structures are defined using construct and are based on the official AmiBroker C++ SDK. This project is not an official AmiBroker API and comes without any warranty. Improvement requests are welcome, but for production usage the official quote downloader might still be the better choice.

Unofficial Resource

A community-written overview is available at deepwiki.com/F2011B/ami2py. This guide is unofficial and may become outdated.

Important

On Windows there are reserved file names such as CON. If such a file is opened and data is written, the output is printed to the console. Therefore the command add_new_symbol renames symbols like CON.DE to C_O_N.DE by default.

Python Package API

The Python module allows you to create new databases and read or append quote information. ami2py exposes dataclasses for the master file and individual symbol records and supports a fast facade for large data volumes.

Examples

Creating a database from scratch and adding symbol data:

from ami2py import AmiDataBase, SymbolEntry

db = AmiDataBase(db_folder)
db.add_symbol("AAPL")
db.append_symbol_entry(
    "AAPL",
    SymbolEntry(
        Close=200.0,
        High=230.0,
        Low=190.0,
        Open=210.0,
        Volume=200003122.0,
        Month=12,
        Year=2020,
        Day=17,
    ),
)
db.write_database()

Read a list of symbols stored in the database:

symbols = db.get_symbols()
print(symbols)

Get values for a symbol in a pandas compatible dictionary format:

db.get_dict_for_symbol("SPCE")

get_dict_for_symbol caches the returned data. Pass force_refresh=True to reload the latest values from disk:

db.get_dict_for_symbol("SPCE", force_refresh=True)

Reading index data:

db.get_dict_for_symbol("^GDAXI")

Using a list container facade for fast reading of symbol data:

data = db.get_fast_symbol_data("SPCE")
newslice = data[0:10]

Updating a database from Yahoo:

python scripts/update_yahoo_db.py /path/to/db

RUST Backend

The parsing logic can optionally be executed using a Rust implementation. Set the environment variable AMI2PY_USE_RUST=1 to activate it (requires the rust_bitparser extension to be built). The included run_tests.sh script builds the extension automatically. Manual build is possible with:

cargo build --manifest-path rust/rust_bitparser/Cargo.toml --release --offline

ami_cli

The rust/ami_cli directory contains a small command line tool written in Rust that wraps the Python API via PyO3. Build it using cargo build --manifest-path rust/ami_cli/Cargo.toml --release.

Examples

ami_cli <command> [args]
  create <db_path> <symbol1> [symbol2 ...]
  add-symbol <db_path> <symbol1> [symbol2 ...]
  list-symbols <db_path>
  list-quotes <db_path> <symbol> [start YYYY-MM-DD end YYYY-MM-DD]
  add-quotes <db_path> <symbol> <csv_file>

On Windows use scripts\build_cli_windows.bat to compile ami_cli.exe. The executable will be found under rust\ami_cli\target\release\ami_cli.exe.

Architecture Overview

The ami2py library is used to read from and write to AmiBroker databases with Python. Important parts of the code base are:

• ami_bitstructs.py – bit structures for individual data entries
• ami_construct.py – assembled binary structures
• ami_dataclasses.py – dataclasses for symbol and master data
• ami_database.py – central AmiDataBase class
• ami_reader.py – reads existing databases
• ami_symbol_facade.py – fast access to symbol data
• consts.py – various constants
• tests/ – unit tests and test data

Key Points

1. Construct structures – the binary formats are defined using construct.
2. Dataclasses – Python dataclasses exist for symbol data and the master file with validation.
3. Fast data access – AmiSymbolDataFacade allows slicing and appending at the binary level to handle large data volumes efficiently.
4. Database folder layout – AmiDbFolderLayout specifies in which subfolder symbols are stored (e.g. a/AAPL).

FAQs

This section collects frequently asked questions about ami2py.